Using the MCP Connector with Uhuu

The Uhuu Document Automation Platform now supports the Model Context Protocol (MCP), enabling developers and operational teams to automate templates, documents, and workflows directly from AI-powered environments like ChatGPT, Claude Desktop, Windsurf, Cursor, and any MCP-compatible IDE.

The Uhuu MCP connector wraps Uhuu's core API capabilities into structured MCP tools that AI assistants can call automatically. This allows you to generate documents, manage templates, and retrieve workspace information in-context, without manually writing integration code.

This guide explains what the connector provides, how to configure it, and how teams can use Uhuu + MCP to automate document workflows end‑to‑end.

1. Overview

The MCP server is available at:

https://mcp.uhuu.io

It exposes tools representing Uhuu's document and workflow automation features:

uhuu_get_team — get team information
uhuu_list_workspaces — list all available workspaces
uhuu_get_workspace — get workspace details
uhuu_get_template — get template details including base_data structure
uhuu_generate_document — generate a document from a template
uhuu_get_document — get document details
uhuu_get_template_documents — list all documents for a specific template

All tools support natural language invocation from your AI assistant.

Example:

"Generate a document from the invoice template with today's order data."

2. Why Use MCP with Uhuu

MCP makes Uhuu an AI-native automation layer for documents, letting assistants run your workflow—no scripting needed.

Benefits

Generate documents instantly with natural-language prompts.
Manage templates and documents through AI assistants.
Access workspace information and template structures.
Streamline document workflows without writing code.

Perfect for High-Volume Document Tasks

Legal, HR, finance, sales, operations, and support.

Uhuu + MCP automates repetitive work and enables true AI-powered document handling.

3. How Non‑Technical Teams Can Use MCP

Even without writing code, teams can instruct an AI assistant to operate Uhuu:

"Generate an NDA document from template ID 123 for this partner."
"Get the template details for template ID 456 to see what data structure it expects."
"List all documents generated from template ID 789."
"Get workspace information for workspace ID 101."

AI becomes the interface for interacting with your workspace, templates, and automations.

4. Connecting to the MCP Server

For Cursor or Windsurf

Add an entry to your MCP configuration:

Cursor (~/.cursor/mcp.json):

{
  "uhuu": {
    "url": "https://mcp.uhuu.io"
  }
}

Windsurf (~/.codeium/windsurf/mcp_config.json):

{
  "uhuu": {
    "url": "https://mcp.uhuu.io"
  }
}

For Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "uhuu": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://mcp.uhuu.io"
      ]
    }
  }
}

Authentication

The MCP server supports multiple authentication methods:

Per-tool API key: Pass api_key parameter in each tool call
Authorization header: Include Authorization: Bearer YOUR_TOKEN in requests
Token management UI: Visit /token on the MCP server to store your token
Environment variable: Set UHUU_API_KEY (for server-side defaults)

Once configured, your assistant will load:

Uhuu MCP tool definitions
Template and document management capabilities
Workspace and team information access

You can then issue natural language commands immediately.

5. Working with Templates

To work with templates, you'll typically:

List workspaces using uhuu_list_workspaces to see available workspaces
Get workspace details using uhuu_get_workspace to see templates in a workspace
Get template details using uhuu_get_template to see the template structure and base_data format

The uhuu_get_template tool is especially important because it returns the base_data field, which shows the expected data structure and example values for generating documents.

Example: Get template details

{
  "template_id": "123",
  "api_key": "YOUR_API_KEY"
}

6. Generating Documents

Generate a document from a template using uhuu_generate_document. Important: Always call uhuu_get_template first to check the template's base_data structure, which shows the expected data format.

Example:

{
  "template_id": "123",
  "body": {
    "customer": "Acme AG",
    "date": "2025-11-12",
    "items": [
      { "label": "Design Services", "amount": 2400 }
    ]
  },
  "api_key": "YOUR_API_KEY"
}

The body parameter should match the structure shown in the template's base_data field.

7. When to Use MCP vs API

Use MCP for:

AI‑assisted automation from within AI environments
Interactive development inside IDEs like Cursor, Windsurf
Natural language document generation
Exploring templates and workspace structures
Rapid prototyping with AI assistance

Use the API for:

Backend integrations and server-side automation
Large‑scale automated systems
Enterprise-level document pipelines
Custom integrations requiring direct API control

MCP is Uhuu's AI-native layer, designed for high‑context, AI-driven productivity within development environments.

Summary

The Uhuu MCP Connector brings the full power of Uhuu’s document automation platform directly into AI-driven tools and editors.

You can:

Get team and workspace information
Retrieve template details and structures
Generate documents from templates with structured data
Access and manage documents
Build document workflows using natural language

Uhuu + MCP transforms your document operations into a fully automated, AI‑powered system.

MCP Server: Build AI-Native Document Automation Workflows with Uhuu